Community-acquired pneumonia (CAP) is the most common serious bacterial infection in children and a leading cause of pediatric hospitalization. Despite its significance, the decision to hospitalize a child with CAP is currently made using clinical factors, imaging and tests that do not accurately or reliably assess disease risk. The result is substantial variation in hospitalization rates and clinical management. Three newer biomarkers - procalcitonin, proadrenomedullin, and pneumococcal bacterial load - have demonstrated potential to improve severity prediction in CAP. Yet, there are currently no evidence-based clinical prediction rules to gauge disease severity or predict clinical outcomes among children with CAP. The extensive variation in care, in addition to the lack of evidence-based decision aids, highlights the critical need for an improved understanding of disease severity and tools to guide management for pediatric CAP. The proposed research will address this important knowledge and practice gap by pursuing the following specific aims: (1) to determine the association between host and pathogen biomarkers and clinical outcomes, including need for hospitalization, in children with CAP; and (2) to develop and internally validate a clinical prediction rule for need for hospitalization of children with CAP. To achieve these aims, the candidate, Todd Florin, MD, MSCE, will leverage an existing study infrastructure developed in conjunction with his mentors during his KL2 award and that has been enrolling children with CAP since July 2013. The robust infrastructure, the primary data source for this K23, includes an active enrollment mechanism in the Cincinnati Children's Hospital Medical Center (CCHMC) ED and the development of a biorepository of specimens from children with CAP. As a pediatric emergency medicine physician, Dr. Florin is uniquely positioned to accomplish the proposed K23 research and training aims. Dr. Florin's long-term goal is to become an expert in clinical and translational research methods to improve the diagnosis and outcomes of pediatric CAP and other common, serious pediatric infections in the acute care setting. In order to expand his research training to include advanced statistical methods in predictive modeling, molecular epidemiology, biomarker discovery, and multicenter pediatric emergency medicine research, Dr. Florin has developed a detailed career development plan consisting of mentorship, didactic coursework, hands-on laboratory experience and completion of the proposed research. He has assembled a multidisciplinary team of experienced mentors with expertise in pediatric pneumonia, infectious diseases research, biomarker discovery, predictive modeling and multicenter emergency medicine research. Dr. Florin will meet regularly with his mentors individually, and quarterly as a full committee, to ensure completion of all research and training benchmarks. The institutional commitment to young researchers at CCHMC, one of the largest pediatric research institutions in the United States, provides an ideal collaborative training environment by providing the strong mentorship, formal instruction, statistical database support, specialized laboratories, and practical experiences necessary to develop into an independent clinical investigator. This proposal will allow Dr. Florin to make significant contributions to pediatric infectious diseases research in the ED as an independent physician-scientist and prepare him for future R01-funded projects focused on use of biomarkers to improve pneumonia diagnosis and management in children.